There are many instances when it is important to know whether, or to what extent, ice has formed on a given surface. The presence of ice on the wings, control surfaces or fuselage of aircraft is known to be potentially life threatening, but there are many other instances (e.g., cooling towers, aerials, refrigerator elements, bridges and roadways and the like) where ice detection is also important. While various proposals have been made in the past for devices capable of detecting ice formation and indicating when ice has formed to an impermissible degree, in general those prior art devices suffer from complexity, expensiveness and unreliability. In many instances they are also undesirable for use with aircraft because they normally project out from the aircraft surface where ice detection is to take place and therefore disturb the aerodynamic efficiency of the craft. For example, the device disclosed in Bullen et al. U.S. Pat. No. 2,803,813 of Aug. 20, 1957 utilizes an element which normally projects beyond the aircraft surface so that if icing conditions prevail ice can form on that object, as a result of which when the object is pulled back into the aircraft its retrograde motion will be impeded by the ice and thus will cause a part within the housing to move and actuate an alarm. In addition to the fact that the part must project out from the airplane for an appreciable period of time in order to give ice the opportunity to form on it if weather conditions are appropriate, the presence of foreign particles other than ice on the exterior of the object will cause false alarms, thus making it unreliable.